1. Field of the Invention
This invention relates to the storage of information on magnetic media, and more particularly to storage of information on rotating magnetic media, for example rotating disks in a disk drive.
2. Description of the Prior Art
Disk drives in the prior art have utilized circular tracks for recording of information on disks having a magnetic coating on the surface. When a plurality of disks are utilized, an assemblage of sequentially or simultaneously addressable tracks is known as a cylinder. It is a goal of disk drive designers to provide as much storage capacity as possible. For a given physical structure, this has typically been accomplished by either increasing the amount of data recorded on a track, or increasing the number of tracks. Disk drives of the prior art design have always used a fixed number of tracks (and cylinders) to obtain a maximum storage capacity.
In the prior art, the surfaces of each disk have been divided into recording zones and zone bit recording techniques utilized such that the zones on each disk surface align vertically. For example, U.S. Pat. No. 4,799,112 to Bremmer et al. describes what is known in the industry as zone bit recording. In this technique, zones are defined on the disk surface and the frequency of recording within a zone is maintained constant. Zone density recording is a piece-wise linear implementation of constant density recording. A specified read/write frequency is different, however, from one zone to another. Referring to FIG. 1 of this application, a highly simplified drawing of a pair of disks 1 and 2 in a disk stack (shown as a cross-sectional-view of these disks) illustrates the recording zones on the disks defined as Z1, Z2 and Z3. As illustrated in FIG. 1 by the dashed lines, the zones on the two-disk stack are vertically aligned and the boundaries of the zones defined radially by the distances indicated by R1-R4 based on the distance from center C of the disk. It will be noted that in the stack the centerline C of the disks are aligned. The read/write frequency used in the corresponding zone of each disk is the same. Thus the read/write frequency in zone Z1 of disk 1 is the same as that used zone Z1 of disk 2 and so on. The recording frequency utilized within each of the zones is determined at the design stage based on various factors, including an expected nominal head read/write performance for the heads to be used in the drive. Based on the expected performance level of read/write heads, a recording frequency was set within each zone to deliver the desired storage capacity for the drive and the number of tracks for each zone was set by other considerations to be discussed later.
The layout of the zones was also based on a consideration of the physical dimensions of the drive in terms of stroke (which will be described in more detail below), the worst case head performance, flying height performance, zone efficiency and the desired yield of drives based on their performance after assembly. With regard to head performance, that is defined in the art on the basis of head performance at a given frequency in terms of the offtrack bit error rate. The offtrack bit error rate is defined as the number of bits transferred per bit in error when data is read at a predetermined offset, in the case described herein at a position 10% off track. Referring to FIG. 2, a graphical illustration is provided of the performance curve of a typical head. The read/write frequency (which is the NRZ frequency) is plotted along the X axis, and on the Y axis the log of the offtrack bit error rate, abbreviated herein as LOBER. In FIG. 2, f.sub.R is the typical average operating frequency of a head, which may be for example about 20 MHz. The disk drive designer defines a minimum offtrack bit error rate threshold, which in FIG. 2 is indicated by the dashed line TH. In the plot illustrated in FIG. 2, the minimum acceptable log offtrack bit error rate TH is equal to 6 (i.e. one bit in error for 10.sup.6 bits transferred, this number being used for illustrative purposes. The diagonal line indicated by reference character 7 represents the performance curve of a nominal head. This curve could move up or down depending on the head design. In the prior art, after assembly of the disk drive the performance of each head with its respective surface was measured to determine if it met the minimum performance standard TH. If any of the heads in the stack of disks failed to reach at least a minimum TH, then the drive was considered unacceptable and was not shipped unless it could be reworked to make it meet the minimum performance standard. Rework in this scenario involved replacing the bad head(s) and/or disk, rewriting the servo information and retesting the drive.
An illustrative, hypothetical example of this test is illustrated in FIG. 3 where the performance of heads HD #1, HD #2, HD #3, and HD #4 are plotted at the typical average operating frequency f.sub.R. Head HD #1 has a log offtrack bit error rate (LOBER) equal to about 5.5 at f.sub.R, while head HD #2 is approximately 6.5, HD #3 about 7.9 and HD #4 approximately 8.5. Following prior art test constraints, since head HD #1 is below the minimum acceptable threshold, the drive failed and would not be usable under traditional formatting conditions.
A second design criterion used by prior art designers was stroke for the heads, the radial travel expected for the read/write heads. One criterion used was worst case stroke, and another single sided detect stroke. Typically one or the other of these values was used in testing of the assembled drives and the drive evaluated based on the measured stroke of the drive after assembly. If the actual stroke was not up to the specification, then the desired number of tracks could not be accessed, the drive was rejected. Commonly, rework would be attempted, which involved replacing or adjusting the crash stops, then rewriting the servo information and retesting the drive.
Referring to FIG. 4A, a graphical illustration is provided of two prior art methods used for specifying the minimal acceptable stroke under the worst case stroke analysis, and secondly under the single sided detect analysis. The bell curve at the OCS and the ICS indicate the expected mechanical tolerance distribution of the location of the outer crash stop and inner crash stop, respectively, on a drive based on the three-sigma (3.sigma.) tolerance values. In FIG. 4A, the vertical line in the center of the ICS and OCS mechanical distribution curves indicates the mean dimension. In the prior art, servo data is written on the disks before the crash stops are placed in the drive. If the position variance of both the inner and outer crash. stop is .sigma..sup.2, then the average position lost (i.e. usable position lost for recording data) on a drive where all tracks are written before the crash stop positions are detected is 6.sigma., the worst case stroke (w.c. stroke) indicated in FIG. 4A. On prior art drives which detect one crash stop and then a write tracks until the second crash stop is detected, the average position lost is: ##EQU1##
Thus it will be appreciated that in both prior art methods since assumptions are made about the usable data area based on the above techniques, there is a loss of available position for storing information.
In prior art products, zone tables were established for a disk drive model at the design stage, with the zone boundaries, the frequency for each zone and the number of tracks to be used being specified. This resulted in zone alignments such as those illustrated in FIG. 1. Although not shown in FIG. 1, if the lower surfaces of disk 1 and 2 were used, as would be typical, the zone boundaries and read/write frequency within the boundaries, would be the same for the upper and lower surfaces of all of the disks. After gaining experience with the drive model based on production, the zone tables were sometimes changed for all drives in subsequent production runs, however for all drives in the subsequent production runs the zone boundaries and the read/write frequency for the zones both remained vertically aligned, again as illustrated in FIG. 1. If it was determined that additional tracks could be used, that may have also been factored in when establishing zone boundaries.
In addition to the foregoing criteria used in the prior art in disk drive design, the number of tracks per inch, also known as the track pitch, was established based on the expected track width of a signal written on a surface of a magnetic disk considering the nominal performance of heads, also referred to as read/write transducers, of the type being utilized in the drive. If the actual width of tracks written by a head was less than the expected/nominal width, then valuable recording space was lost since the tracks could have been spaced more closely without track-to-track interference.